Data analytics is something that you read and hear about a lot these days. It’s been touted as the science and technology that can help organizations reduce costs, increase efficiency, identify new opportunities and increase their competitive advantage.
Data analysts uses mathematics, statistics, and sometimes machine learning, to find meaningful patterns across registered data and turn them into meaningful results and actionable insights. While the concept of analytics isn’t new, recent advances in storage and computation technologies are enabling us to store and process data in huge volumes, and are taking data analytics to the next step, opening up possibilities that weren’t conceivable before.
Applying analytics to the data we gather will help us answer questions about how and why things happen, what might happen in the future and what decisions and actions need to be taken. Data analytics can be broken down to the three fields of descriptive, predictive and prescriptive. Here’s what you need to know about each.
Descriptive analytics
This is the vanilla of analytics and it relies on gathering facts and summarizing what has happened in the past by browsing and classifying data. This model helps you understand the how and why of your operations. Almost all management software such as CRM, ERP, ecommerce and social media networks have a form of descriptive analytics packed in. If you’ve been blogging or managing a site, you’ve definitely had a taste of descriptive analytics.
For instance, the Stats feature that is included in all blogs and sites powered by WordPress.com (including mine) is a good example of a descriptive analytics engine. Here, you can see how many people visited your site, how many pages were viewed, which pages were viewed, from which sites users were referred to your site, the geographical locations, etc.
The Twitter analytics page is another example of descriptive analytics used in social media.
Descriptive analytics engines usually use charts, graphs and tables to display data.
While descriptive analytics won’t give you forecasts about what will happen in the future, there are lots of things a savvy user can do if you put it to good use. For instance, with descriptive analytics, you can segment and categorize your data to achieve better understanding.
One of the best uses of descriptive analytics is A/B testing, where you compare user or customer reactions to different versions of the same thing. For instance, when you want to decide where to place a specific button on your webpage, a little programming and the proper use of analytics tools can give you precise data about which one is generating more clicks.
Predictive analytics
Predictive analytics turns data into actionable information. As the name implies, predictive analytics will help you forecast future probabilities with an acceptable level of reliability. This is done by analyzing past data, formulating a statistical model and running algorithms to predict behavior and events. Machine learning and additional data is used to revise and enhance the model.
Predictive analytics has many great uses. In the sales and marketing domain, it is used to optimize customer relationship management and predict the next move of a potential customer. In the insurance industry, it is used to take into account potential driving safety predictors such as age, gender, and driving record when issuing car insurance policies.
Predictive analytics is also being used in the cybersecurity industry to discover data breaches before they happen.
Predictive analytics can also be used to fill in the gaps and predict data that is not there. The most prevalent use case in this regard is sentiment scoring.
Prescriptive analytics
Prescriptive analytics is the evolution of its predecessors. In addition to gleaning insights and making predictions, prescriptive analytics will answer the question of “what to do?” and will offer suggestions on what specific actions can give you the most value for the specific endeavor you’re engaged in.
Prescriptive analytics helps you make decisions that will take advantage of future opportunities or mitigate future risks. Prescriptive analytics can also guide you on the possible implications of future decisions. By continually ingesting and processing new data, prescriptive analytics improves its prediction accuracy and provides better decision suggestions.
For instance, healthcare planning can gauge the cost-benefits of creating new facilities, increasing or decreasing production of medicine, etc. based on historical data, predictions and suggestions made by prescriptive analytics engines.
Prescriptive analytics relies on big data, business rules and mathematical sciences to perform its functionalities. Data can be structured or unstructured, and it can come from the organization’s internal sources or collected from social media and the web. Business rules can be boundaries, constraints, preferences, policies and best practices. The mathematics behind prescriptive analytics include applied statistics, machine learning algorithms and natural language processing.